RANO 2.0 — Glioma Response Assessment MRI

MRIninja Knowledge Base | Focus / Deep Dive Page Version 1.0 — April 2026 Companion pages: MRI Brain Generic Standard Protocol; DSC Perfusion MRI Series (Parts I–III)

1. Why RANO 2.0 Matters to the Radiologist

Response Assessment in Neuro-Oncology (RANO) 2.0, published in the Journal of Clinical Oncology in September 2023 [1], is the current authoritative framework for evaluating treatment response in adult gliomas in clinical trials. For the practising radiologist, understanding RANO 2.0 is not merely an academic obligation — it directly determines how images are described, which measurements are reported, and how the words in a report influence treatment decisions in neuro-oncology.

RANO 2.0 supersedes four previous criteria that applied to different glioma populations and treatment contexts:

• RANO-HGG (2010): for high-grade gliomas in standard chemoradiation trials
• RANO-LGG (2011): for low-grade, primarily non-enhancing gliomas
• mRANO (modified RANO): a practical modification adopted by many centres addressing the pseudoprogression window
• iRANO (immunotherapy RANO): a variant for immunotherapy trials

RANO 2.0 unifies all glioma subtypes and treatment modalities into a single framework, aligning with the WHO 2021 CNS tumour classification that defines gliomas molecularly rather than purely histologically [1, 2].

The key conceptual shift in RANO 2.0 relative to all previous RANO versions is the recognition that the imaging baseline matters more than the imaging change. By moving the baseline reference scan from the post-surgical scan to the post-radiotherapy scan, RANO 2.0 directly addresses the pseudoprogression problem that made RANO-HGG unreliable in the first 12 post-radiation weeks.

2. The RANO 2.0 Framework: Core Principles

2.1 Three Imaging-Based Tumour Categories

RANO 2.0 classifies gliomas into three imaging-based categories that determine which MRI measurements are used and which response criteria apply. This classification is driven by the molecular type of the tumour and its dominant imaging appearance:

Category 1 — Enhancing tumour (primarily contrast-enhancing disease): Applied to IDH-wildtype glioblastoma (the dominant clinical scenario). Measurements are based on post-contrast T1-weighted enhancement. T2/FLAIR abnormalities are no longer measured as part of the primary response assessment for this category — a major change from RANO-HGG, where FLAIR progression could trigger disease progression designation even without enhancement change [1, 3]. For sequence-level protocol optimisation, vendor terminology and artefact management, see the dedicated MRIninja page FLAIR Sequence.

Category 2 — Non-enhancing tumour (primarily non-enhancing disease): Applied to IDH-mutant, predominantly non-enhancing gliomas (most Grade 2–3 IDH-mutant astrocytomas and oligodendrogliomas). Measurements are based on T2/FLAIR abnormality area or volume. Post-contrast T1-weighted enhancement is monitored for new development (which would upgrade the response category).

Category 3 — Mixed tumour (both enhancing and non-enhancing components): Applied to tumours with measurable components in both modalities — including IDH-wildtype GBM on anti-angiogenic therapy (where enhancement decreases but non-enhancing T2/FLAIR may progress), high-grade IDH-mutant gliomas with mixed morphology, and diffuse midline gliomas H3K27-altered. Both enhancing and non-enhancing components are assessed, though for progressive disease determination, measuring only the enhancing lesion is also allowed [1, 4].

2.2 The New Baseline: Post-Radiotherapy MRI

In the newly diagnosed setting, RANO 2.0 mandates that the post-radiotherapy MRI — acquired 21–35 days after completion of radiation therapy — serves as the baseline reference scan for all subsequent response assessments. This replaces the prior convention of using the post-surgical, pre-radiation MRI as baseline [1, 2, 5].

The rationale is compelling and data-driven: treatment-related changes (pseudoprogression) are most common within the first 12 weeks post-radiation, occurring in 20–30% of all patients and up to 40–50% of MGMT-methylated GBMs [3]. Using the post-surgical scan as baseline means that pseudoprogression — which produces an increase in enhancement relative to the post-surgery baseline — would be misclassified as progressive disease (PD) under RANO-HGG. RANO 2.0 avoids this by resetting the baseline to the post-radiation state, after the acute treatment-related changes have partially declared themselves.

Practical consequence for the radiologist: every glioma follow-up report in the newly diagnosed setting must identify which scan is the RANO 2.0 baseline and provide measurements referenced to that baseline, not to the pre-radiation or post-surgical scan.

The post-radiation baseline scan timing (21–35 days after completion of RT) is standardised to allow sufficient time for acute treatment effects to stabilise while avoiding the period of maximal pseudoprogression. For recurrent glioma trials, the baseline scan is the most recent pre-treatment scan.

3. MRI Sequences Required: The Brain Tumour Imaging Protocol (BTIP)

RANO 2.0 mandates adherence to the Brain Tumour Imaging Protocol (BTIP), developed in 2015 [6] and updated in the 2023 integration guidelines [7]. The BTIP specifies exact sequence requirements and acquisition parameters designed to ensure reproducible, measurement-ready image quality across time points and centres.

3.1 BTIP Mandatory Sequences

Advanced sequences (DSC perfusion, MR spectroscopy, DCE) are additive to the BTIP but not mandated by RANO 2.0 itself. DSC perfusion is endorsed as adjunctive for pseudoprogression assessment but not yet formally incorporated into RANO 2.0 response criteria [1].

3.2 The Parameter-Matched Pre/Post Contrast T1 Requirement: Why It Is Not Optional

The most technically demanding BTIP requirement — and the most frequently violated in routine practice — is the parameter matching between pre- and post-contrast 3D T1-weighted acquisitions. The pre-contrast T1 must be acquired with identical sequence parameters (matrix, FOV, slice thickness, TR, TE, TI, flip angle) and identical geometry (same slice prescription, same orientation, same FOV position) as the post-contrast T1.

This requirement exists because the BTIP standard is designed to support T1 subtraction maps (post minus pre contrast), which provide far higher sensitivity for subtle enhancement than visual comparison of pre and post images alone. T1 subtraction eliminates background T1-bright structures (haemorrhage, fat, mineralization, enhancing vessels) and reveals only genuine gadolinium-induced T1 shortening. If the pre and post T1 acquisitions are not parameter-matched, any geometric mismatch between the two images will produce subtraction artefacts that simulate or obscure enhancement — making the subtraction map unreliable [6, 7].

For the MRI technologist: BTIP compliance requires that the pre-contrast T1 sequence is not a quick scout or a lower-resolution "for comparison" acquisition. It must be a full-quality 3D isotropic T1 with identical parameters to the post-contrast acquisition. The pre-contrast T1 must be acquired before gadolinium injection and the post-contrast T1 must be acquired in the same position without patient movement. Any patient repositioning between the two acquisitions degrades subtraction quality.

3.2a Why IR-GRE (MPRAGE/BRAVO/TFE) and Not 3D TSE T1?

A legitimate technical question arises from the BTIP specification: high-quality 3D T1-weighted TSE sequences (SPACE T1, CUBE T1, VISTA T1) are well established in many implementations and offer excellent grey-white matter contrast. Why, then, does RANO 2.0 mandate the inversion recovery gradient echo (IR-GRE) family rather than 3D TSE T1? For sequence-level protocol optimisation, vendor terminology and artefact management, see the dedicated MRIninja page Gradient Echo (GRE/FLASH) Sequence.

The reasons are technical, practical, and evidence-based — and are not reducible to institutional diffusion alone, although that factor does play a secondary role.

Subtraction reliability and T1 contrast mechanism: the primary purpose of the BTIP T1 sequence is subtraction imaging (post minus pre) for enhancement detection, not grey-white matter differentiation. IR-GRE sequences (MPRAGE, BRAVO, TFE-IR) generate T1 contrast through inversion recovery preparation combined with a rapid gradient echo readout, producing a well-characterised and highly reproducible T1 weighting that behaves predictably with gadolinium. The T1 shortening from gadolinium in an IR-GRE sequence produces a directly proportional and spatially stable signal increase that subtraction algorithms exploit reliably. 3D TSE T1 sequences generate T1 contrast through a different mechanism — modulation of echo train refocusing — which introduces T2 contamination proportional to ETL length, magnetisation transfer effects from the multiple refocusing pulses, and sensitivity to motion between echoes within the echo train. These factors make subtracted images from 3D TSE T1 more prone to artefactual signal differences that are not enhancement-related. For sequence-level protocol optimisation, vendor terminology and artefact management, see the dedicated MRIninja page MPRAGE / 3D T1 Magnetisation-Prepared GRE Sequence.

Sensitivity to subtle gadolinium enhancement: while high-quality 3D TSE T1 sequences produce visually impressive anatomical images, their T1 sensitivity per unit gadolinium concentration is lower than IR-GRE sequences at the standard clinical dose of 0.1 mmol/kg. The inversion recovery preparation pulse in MPRAGE/BRAVO selectively suppresses background signal at the TI, amplifying the relative signal increase from gadolinium-enhanced tissue. For the specific RANO task — detecting small foci of new enhancement or measuring the precise enhancing lesion boundary — IR-GRE provides more robust and reproducible enhancement conspicuity across vendors and field strengths.

Multicentre reproducibility: RANO 2.0 is designed for clinical trials across multiple institutions, vendors, and field strengths. IR-GRE sequences (MPRAGE on Siemens, BRAVO on GE, TFE-IR on Philips, QuickBrain3D on Canon) are implemented on essentially every modern 1.5T and 3T scanner with well-characterised and documented parameters, validated across multiple prospective brain tumour studies. 3D TSE T1 sequences show substantially greater vendor-to-vendor parameter variation — ETL, bandwidth, refocusing pulse design — making their contrast characteristics less reproducible across sites. This vendor heterogeneity would undermine the cross-site reproducibility that multicentre trials require.

Isotropic volumetric capability: IR-GRE sequences are designed from the outset for isotropic 1 mm full-brain acquisition within 5–8 minutes — the volumetric measurement standard supported by RANO 2.0. Their gradient echo readout is inherently efficient for this coverage. 3D TSE T1 at equivalent isotropic resolution typically requires longer acquisition times and produces ETL-related blurring of fine structures that reduces measurement precision at lesion margins.

The institutional diffusion argument: the intuition that MPRAGE-type sequences are more widely deployed and more consistently optimised than high-quality 3D TSE T1 across clinical sites is correct — and not trivial. A trial that mandates a sequence achievable at 95% of participating centres is more practical than one requiring specific vendor expertise to optimise. However, this is a secondary consideration that reinforces the primary technical rationale rather than replacing it.

In summary: IR-GRE is mandated not because 3D TSE T1 is inferior as an anatomical sequence, but because for the specific RANO task of reproducible subtraction-based enhancement measurement across heterogeneous multicentre settings, IR-GRE provides more predictable gadolinium sensitivity, more reliable subtraction quality, and greater cross-vendor standardisation.

3.3 Sequence Acquisition Order for RANO 2.0 Compliance

The sequence order within the brain MRI protocol is not arbitrary in the BTIP context:

1. DWI (pre-contrast mandatory — gadolinium affects ADC values)
2. Pre-contrast 3D T1-weighted (IR-GRE) — the mandatory BTIP pre-contrast sequence
3. Pre-contrast 2D T2-FLAIR (before gadolinium)
4. Gadolinium injection (0.1 mmol/kg macrocyclic agent, power injector)
5. 2D axial T2-weighted TSE (acquired after contrast, before post-contrast T1 — controls contrast timing and acts as a spacer for consistent enhancement timing)
6. Optional: DSC perfusion (using the GBCA already injected, or as separate bolus per DSC protocol)
7. Post-contrast 3D T1-weighted (IR-GRE) — parameter-matched to pre-contrast acquisition

If DSC perfusion is included, the T2-weighted spacer sequence may be replaced by the DSC acquisition itself, providing equivalent timing control. The post-contrast FLAIR, if acquired, should follow the post-contrast T1 [7].

4. Measurement Methodology

4.1 Two-Dimensional (Bi-Planar) Measurements — The Primary Method

RANO 2.0, like its predecessors, uses 2D bi-planar measurements as the primary measurement method: the product of the two largest perpendicular diameters of the enhancing tumour (or T2/FLAIR abnormality for non-enhancing tumours), measured on a single slice [1].

Measurement rules for 2D:

• Measure the largest single cross-sectional area of the contrast-enhancing lesion on post-contrast T1
• The measurement must be made on the same slice plane across all timepoints (axial preferred for most supratentorial lesions)
• The two diameters must be perpendicular to each other
• Each diameter must be ≥ 10 mm to qualify as a measurable lesion (the minimum size threshold has not changed from RANO-HGG)
• Report the sum of the product of perpendicular diameters (SPD) when multiple target lesions are present

Lesions that do not meet the ≥ 10 × 10 mm threshold are classified as non-measurable. They are still documented and monitored, but they do not contribute to the quantitative SPD calculation used for response classification.

4.2 Volumetric Measurements — The RANO 2.0 Option

RANO 2.0 introduces volumetric measurements as an acceptable alternative to 2D for specific scenarios [1, 5]. Volumetric measurement is:

• Preferred for IDH-mutant non-enhancing gliomas, where the T2/FLAIR lesion grows slowly and diffusely, making 2D measurements less reliable than volume estimation [1]
• Optional for all glioma types — trials may mandate volumetric assessment if the sponsor requires it
• Not universally mandated — the measurement approach must be predefined in the trial protocol and applied consistently throughout the study

Volumetric measurements require semiautomated or automated tumour segmentation software. Manual voxel-by-voxel segmentation is impractical in clinical settings. The reliability of volumetric measurements depends on the segmentation algorithm and its consistency across timepoints — any change in software version or algorithm between scans can introduce measurement artefact that mimics tumour change.

The RANO 2.0 progressive disease threshold using volumetric measurements is ≥ 40% volume increase (compared to ≥ 25% SPD increase with 2D measurements) [1].

4.3 Target Lesions, Non-Target Lesions, and New Lesions

RANO 2.0 maintains the target/non-target/new lesion framework from its predecessors:

Target lesions: up to 5 measurable lesions (≥ 10 × 10 mm) selected at baseline. All target lesions are measured at every timepoint and their SPD summed. Lesions that are chosen as target lesions must be reproducibly measurable — lesions adjacent to surgical cavities, in eloquent cortex with susceptibility artefact, or in the posterior fossa near air-bone interfaces may not qualify.

Non-target lesions: lesions present at baseline that are too small to measure, too irregular, or not selected as target lesions. They are documented as "present" or "absent" and qualitatively assessed as "improved", "stable", or "worsened". Non-target progression alone (≥ 25% area increase making a previously non-measurable lesion measurable, defined as ≥ 10 × 10 mm) can trigger progressive disease if the SPD of the newly measurable lesion plus existing target lesions meets the ≥ 25% threshold [1, 4].

New lesions: any unequivocal new enhancing lesion not present at baseline — regardless of its size — constitutes progressive disease under RANO 2.0. However, a critical modification in RANO 2.0 is that a new lesion observed within 12 weeks of completing radiation therapy is classified as preliminary progressive disease (pPD) rather than confirmed PD, and requires confirmation on a subsequent scan [1].

5. Response Categories: Definitions and MRI Correlates

RANO 2.0 defines five response categories for Category 1 (enhancing) tumours [1]:

Major Response (MaR) is a new category introduced in RANO 2.0, filling the gap between PR (25–50% decrease) and CR (complete disappearance). Its addition recognises that a 50% decrease threshold for PR, which was adopted from solid tumour criteria (RECIST), may be too stringent for brain tumours where complete anatomical disappearance of enhancement is uncommon.

For Category 2 (non-enhancing) tumours, the same response thresholds apply but are measured on T2/FLAIR area or volume, not enhancement.

6. The Pseudoprogression Problem: The 12-Week Rule and Confirmation Scans

6.1 Why the 12-Week Rule Was Introduced

One of the most consequential practical changes in RANO 2.0 is the mandatory 12-week confirmation window for apparent progressive disease in the post-radiation setting. Under RANO-HGG, an increase in enhancement within the radiation field at the first post-radiation imaging time point was classified as PD if it met the ≥ 25% SPD threshold. This led to systematic misclassification of pseudoprogression as true tumour progression — particularly in MGMT-methylated GBM — causing patients to be incorrectly removed from effective therapies.

RANO 2.0 addresses this with a two-level system for the 12-week post-radiation window [1, 5]:

Preliminary Progressive Disease (pPD): assigned when the imaging criteria for PD are met within 12 weeks of completing radiotherapy. This is NOT a confirmed progression. It triggers a mandatory confirmation scan at the next imaging timepoint (typically 4–8 weeks later).

• If the confirmation scan shows further increase: confirmed PD, treatment discontinuation appropriate
• If the confirmation scan shows stabilisation or improvement: reclassified as pseudoprogression (treatment effect); continue current therapy; backdate the progression date if subsequently confirmed

Outside the 12-week window: apparent PD with unequivocal progression outside the radiation field (or occurring > 12 weeks after RT) can be classified as confirmed PD without a mandatory confirmation scan. However, RANO 2.0 still allows (and in many trial settings mandates) a confirmation scan to improve reliability [1].

6.2 What Counts as Outside the Radiation Field

A new lesion is considered "outside the radiation field" when it develops clearly beyond the high-dose radiation zone — defined as beyond the 80% isodose line. This determination requires access to the radiotherapy planning data (isodose lines), which should be available in the patient's electronic record and directly compared to the MRI localisation.

The practical implication for radiologists: in the early post-radiation period, a new enhancing lesion within the radiation field should be reported as "preliminary progressive disease — confirmation required per RANO 2.0 criteria", not as confirmed progression. A new lesion clearly outside the radiation field (e.g., contralateral hemisphere, cerebellum when no infratentorial radiation was given) does not require confirmation and constitutes confirmed PD.

6.3 Corticosteroid Effect on Response Classification

Corticosteroid therapy produces pseudoresponse by reducing vascular permeability and decreasing enhancement, mimicking genuine tumour response. RANO 2.0 maintains the corticosteroid qualification rules from RANO-HGG:

• CR and MaR/PR require that corticosteroids are stable or reduced relative to baseline dose at the time of the response evaluation. A decrease in enhancement driven by a concurrent increase in corticosteroid dose is downgraded to SD — or excluded from counting as response — because the mechanism is pharmacological vascular effect, not tumour regression.
• Conversely, if MRI shows apparent PD but corticosteroids have been decreased since the nadir, the response may be reclassified as SD if the clinical and radiological picture supports it [1, 2].

7. Non-Enhancing Disease in RANO 2.0: The T2/FLAIR Component

7.1 What RANO 2.0 Says About Non-Enhancing Disease

RANO 2.0 makes a significant modification to the role of T2/FLAIR in response assessment for enhancing (Category 1) tumours: T2/FLAIR progression alone no longer constitutes progressive disease in predominantly enhancing glioblastoma [1, 3]. Under RANO-HGG, a significant increase in T2/FLAIR abnormality could be classified as PD even without enhancement change. This was problematic because T2/FLAIR changes are highly non-specific — they are seen with pseudoprogression, radiation-induced white matter changes, vasogenic oedema from steroids, and infiltrative non-enhancing tumour — and did not reliably reflect active tumour.

The modification: for Category 1 (primarily enhancing IDH-wildtype GBM), progression is determined by enhancement measurements only. T2/FLAIR is still monitored and reported, and major unexplained T2/FLAIR increase should be noted, but it does not independently trigger a PD designation.

Exceptions: T2/FLAIR measurements do contribute to PD determination in:

• Category 2 (non-enhancing) tumours — T2/FLAIR is the primary measurement
• Category 3 (mixed) tumours — both enhancement and T2/FLAIR contribute
• Anti-angiogenic therapy: the reduction in enhancement during bevacizumab therapy ("pseudoresponse") may mask T2/FLAIR progression. For patients on anti-angiogenic agents, RANO 2.0 retains T2/FLAIR assessment as a counterweight to the masking of enhancement [1]

7.2 Imaging T2/FLAIR Abnormalities: What to Measure and What to Report

For Category 2 non-enhancing gliomas, T2/FLAIR measurement follows the same 2D or volumetric framework as enhancement measurement for Category 1. The challenges are:

• Lesion boundary definition: T2/FLAIR abnormality in IDH-mutant gliomas fades gradually into normal white matter without a sharp border. The measurement must be performed consistently at a predefined contrast threshold (RANO 2.0 does not specify a threshold but requires consistency across timepoints). Ideally, the same contrast window/level settings and the same measurement plane are used at every timepoint.
• Oedema vs. infiltrative tumour: T2/FLAIR hyperintensity in glioma contains a mixture of tumour infiltration and peritumoral vasogenic oedema. RANO 2.0 measures the composite abnormality, not the infiltrative component alone. This is a known limitation — reducing oedema with steroids will reduce T2/FLAIR signal and could simulate response.
• Treatment effects: radiation-induced changes, temozolomide-related white matter changes, and FLAIR pseudo-worsening from increased blood-brain barrier permeability post-radiation all affect T2/FLAIR in ways that may not reflect tumour change.

7.2.1 The Boundary Problem: Infiltrative Margin vs. Vasogenic Oedema

The most fundamental challenge in T2/FLAIR measurement of Category 2 gliomas is that the T2/FLAIR hyperintense region is not a single tissue type — it is a spatial continuum from solid tumour core to infiltrative margin to pure vasogenic oedema to normal-appearing white matter, without sharp histological or imaging transitions between zones. RANO 2.0 measures the full composite T2/FLAIR abnormality, acknowledging the impossibility of separating these components on standard anatomical sequences alone [1]. Understanding the biology and the available imaging approaches for approaching this boundary is essential for reproducible measurement practice.

7.2.2 Pathological Basis of the Peritumoral Zone in IDH-Mutant Glioma

In IDH-mutant diffuse gliomas, the non-enhancing peritumoral area (NEPA) — defined as the T2/FLAIR hyperintense region surrounding the tumour core — corresponds pathologically to two distinct processes that cannot be separated by conventional MRI sequences alone [18]:

Infiltrative tumour: individual or small clusters of neoplastic cells migrating along white matter tracts, perivascular spaces, and subpial surfaces, without inducing angiogenesis or blood-brain barrier disruption. This zone contains tumour cells at variable density, often indistinguishable from normal brain on post-contrast T1 imaging. Its T2/FLAIR signal is driven by gliosis, increased extracellular water content from disrupted cell membranes, and altered myelin structure.

Vasogenic oedema: fluid accumulation in the extracellular space secondary to blood-brain barrier disruption from tumour-secreted growth factors and inflammatory mediators, without necessarily containing tumour cells. Vasogenic oedema produces higher T2/FLAIR signal intensity than infiltrative tumour in many cases because of the larger volume of free water, but the signal intensity overlap is substantial and cannot be used as a reliable discriminator on FLAIR alone.

IDH-mutant gliomas, compared to IDH-wildtype GBM, tend to produce relatively less vasogenic oedema relative to the volume of infiltrated brain — a consequence of their lower VEGF expression and more intact microvascular architecture [10]. This means that in IDH-mutant diffuse gliomas (the dominant Category 2 population), the T2/FLAIR hyperintensity is proportionally more dominated by infiltrative tumour than in GBM, and the FLAIR boundary is more likely to reflect true tumour extent. However, this is a probabilistic statement, not a reliable distinction at the individual voxel level.

7.2.3 The T2-FLAIR Mismatch Sign: Implications for Boundary Assessment

IDH-mutant, 1p/19q non-codeleted astrocytomas frequently display the T2-FLAIR mismatch sign — a pattern of homogeneous T2 hyperintensity combined with relative FLAIR hypointensity in the tumour core. This sign reflects microcyst formation, high free water content, and specific T1 relaxation characteristics of IDH-mutant astrocytoma tissue and has high specificity (approximately 90%) for IDH-mutant non-codeleted status [12]. Its practical implication for boundary measurement is significant: when the mismatch sign is present, the T2 image may show a sharper tumour core boundary than FLAIR, because the core T2 hyperintensity is distinct from the perilesional FLAIR changes. In this scenario, the T2 image should be consulted alongside FLAIR for boundary definition, and the outer FLAIR boundary — which reflects the broader infiltrative and oedematous zone — may overestimate the solid tumour extent relative to the T2 boundary.

IDH-mutant oligodendrogliomas (1p/19q codeleted) do not show the mismatch sign and typically display concordant T2 and FLAIR hyperintensity throughout their extent, making boundary definition harder on both sequences simultaneously.

7.2.4 Advanced MRI for Boundary Definition: Evidence and Limitations

Multiple advanced MRI parameters have been investigated for their ability to distinguish infiltrative tumour from vasogenic oedema within the peritumoral T2/FLAIR hyperintense zone. None has sufficient prospective validation to be incorporated into RANO 2.0 criteria, but they are used as adjunctive tools in complex cases.

DWI / ADC mapping is the most clinically accessible advanced tool. The rationale: tumour-infiltrated tissue has increased cellularity relative to pure oedema, which restricts water diffusion and produces lower ADC values.

• Pure vasogenic oedema: ADC typically 1.60–1.80 × 10⁻³ mm²/s, reflecting unimpeded water diffusion in a protein-rich extracellular compartment
• Infiltrative tumour margin: ADC typically 1.10–1.40 × 10⁻³ mm²/s, reflecting moderate cellularity restricting diffusion without the complete restriction of densely cellular tumour core
• Solid tumour core (non-enhancing): ADC 0.90–1.20 × 10⁻³ mm²/s depending on grade and molecular type; IDH-mutant gliomas tend to have higher ADC than GBM due to microcyst formation and lower cellularity

A 2025 AJNR study specifically examining the boundary between non-enhancing tumour and vasogenic oedema in GBM documented an optimal ADC cutoff of 1.36 × 10⁻³ mm²/s (sensitivity 0.93–1.0, specificity 1.0) for distinguishing solid non-enhancing glioma from pure vasogenic oedema [9]. However, perilesional non-enhancing T2/FLAIR hyperintensity had ADC values overlapping with vasogenic oedema (1.67 vs 1.74 × 10⁻³ mm²/s; P = 0.32), confirming that ADC alone cannot separate peritumoral infiltration from oedema in the intermediate zone — the most clinically relevant boundary region.

DSC perfusion / rCBV: tumour-infiltrated tissue has elevated rCBV relative to vasogenic oedema, because even low-density infiltrative tumour induces modest microvascular recruitment. Pure vasogenic oedema shows rCBV at or below normal white matter (typically nrCBV 0.3 or less). The same 2025 AJNR study documented that an rCBV cutoff of 0.42 distinguished perilesional tumour infiltration from pure vasogenic oedema with 86% specificity [9]. This is clinically useful as a supplementary tool when DSC is available, though it does not achieve sufficient sensitivity/specificity for voxel-level tumour boundary delineation.

Multiparametric combination: ADC and rCBV together substantially outperform either alone for the vasogenic oedema vs. infiltrative tumour distinction. A voxel or region that shows both reduced ADC (< 1.36 × 10⁻³ mm²/s) and elevated rCBV (nrCBV > 0.42) in the peritumoral zone is more confidently classified as infiltrative tumour than vasogenic oedema. A region with high ADC and low rCBV is more likely to be predominantly oedema. The intermediate zone — high ADC and intermediate rCBV — remains genuinely ambiguous and represents the biological reality of mixed tissue composition [12, 15].

DTI / fibre tractography: directional diffusivity metrics (fractional anisotropy, radial diffusivity) reflect white matter tract integrity within and around the tumour. Tumour infiltration tends to reduce FA more than pure oedema and produces characteristic patterns along fibre tracts. DTI-based boundary assessment has been used in surgical planning contexts but has not been validated for RANO response assessment.

Amino acid PET (FET, FDOPA, MET): uptake of amino acid tracers reflects increased amino acid transport by tumour cells, which is present in infiltrative tumour but not in vasogenic oedema. The tumour-to-background ratio on amino acid PET can identify infiltrative tumour in the T2/FLAIR hyperintense zone with greater specificity than ADC or rCBV. PET RANO 1.0 (Lancet Oncology 2024) provides a parallel framework for PET-based response assessment in glioma [8]. In specialised centres with amino acid PET availability, co-registration of PET with MRI provides the most accurate mapping of the biologically active tumour extent beyond the conventional T2/FLAIR boundary.

7.2.5 Segmentation Criteria Used in Published Clinical Studies

Understanding how published studies have approached T2/FLAIR segmentation in non-enhancing gliomas is essential context for interpreting their results and for designing reproducible institutional workflows. The heterogeneity of methods across the literature is itself the main reason why T2/FLAIR measurement variability remains a fundamental limitation of RANO for Category 2 gliomas [11].

Manual segmentation on T2/FLAIR (the most common approach): The majority of published volumetric studies in IDH-mutant low-grade glioma have used manual segmentation of the T2/FLAIR hyperintense region by trained readers, using software such as ITK-SNAP, 3D Slicer, or BrainLab. A representative protocol used in longitudinal studies [17] segmented the T2/FLAIR hyperintense region slice-by-slice on axial images, including all signal clearly above normal white matter intensity as determined by visual comparison with the contralateral hemisphere. No explicit signal intensity threshold was specified; the definition was fundamentally visual and reader-dependent. Inter-reader reproducibility in this approach is moderate — even among expert neuroradiologists, mean diameter measurements (MTD) for DLGG show variability of 5–15% in published series [17].

The RANO-LGG criteria (2D perpendicular diameters): The original RANO-LGG (2011) [20] and subsequent RANO criteria defined tumour measurement as the product of the two largest perpendicular diameters of the T2/FLAIR hyperintense lesion on a single axial slice. No explicit threshold for signal intensity was defined; the reader identified the largest cross-section and measured within the visible T2/FLAIR signal change. A 2023 retrospective study of 63 LGG patients documented that this approach achieved accuracy of only 57–67% compared to volumetric gold standards, with wide variability even between board-certified neuroradiologists, particularly at lesion boundaries [11]. The main sources of error were: selection of different representative slices, different interpretation of the outer tumour boundary, and inability to capture complex or multilobar tumour morphology.

Semi-automated threshold-based segmentation: Several groups have used signal intensity threshold-based approaches to improve reproducibility. A commonly used method defines the tumour boundary as the 3D region with T2/FLAIR signal intensity above a predefined multiple (typically 1.5–2.0×) of the mean NAWM signal in the contralateral hemisphere. This approach reduces reader-dependent boundary uncertainty but introduces threshold-selection dependence — the resulting volume changes systematically if the threshold is modified, and the optimal threshold is not standardised across the literature. The BTIP consensus [4] does not mandate a specific threshold approach for Category 2 measurement.

Deep learning-based automated segmentation: Automated segmentation of non-enhancing glioma on T2/FLAIR is technically feasible with convolutional neural networks trained on BraTS datasets (HD-GLIO, nnU-Net architectures). However, published studies consistently report that non-enhancing LGG segmentation is more difficult than high-grade enhancing tumour segmentation because of diffuse borders and variable signal intensity at the tumour margin [16]. The EASE study (Frontiers in Medicine 2021) [16] showed a clinical implementation success rate of approximately 74% for individual scans, with failures predominantly at uncertain boundary regions. Automated methods that achieve excellent performance on high-grade glioma (enhancing tumour) may underperform specifically on the non-enhancing margin that is most clinically relevant for Category 2 measurement.

Key parameters reported in published segmentation studies:

7.2.6 Practical Protocol for Reproducible T2/FLAIR Measurement in Category 2 Gliomas

Given the absence of a mandated threshold in RANO 2.0, the following practical approach is recommended for clinical and trial use to maximise reproducibility:

1. Standardise the FLAIR sequence parameters across all serial examinations (3D FLAIR preferred; same TR/TE/TI, same geometric prescription, same slice thickness). Any change in FLAIR acquisition parameters produces apparent signal intensity changes that are artefactual rather than biological.
2. Display standardisation: at each measurement timepoint, set the image window/level to the same values as the baseline examination. The most reliable approach is to calibrate the window using the contralateral NAWM mean signal as the lower reference and the CSF as the upper reference, holding these constant.
3. Use the same axial plane as defined at baseline, unless the tumour has grown to require a different representative slice. Document the slice number and coordinates if using volumetric imaging.
4. Outer boundary decision rule: measure to the outermost extent of clearly elevated T2/FLAIR signal relative to the contralateral hemisphere. Do not attempt to exclude peritumoral oedema from the measurement — RANO 2.0 measures the composite abnormality.
5. For ambiguous boundaries: co-register the ADC map and, where available, the rCBV map to assist with boundary decision. Regions with ADC > 1.50 × 10⁻³ mm²/s and rCBV clearly below normal white matter (nrCBV < 0.5) at the tumour margin are predominantly oedema and may represent the outer limit of the measured region. This is not a mandatory criterion but a reproducible clinical practice aid.
6. Document explicitly in every report whether the T2/FLAIR measurement includes the full composite signal (standard RANO approach) or excludes a clearly defined oedema rim (non-standard, must be declared and maintained throughout the study).
7. For corticosteroid changes: if the corticosteroid dose has been modified between examinations, document this prominently. A decrease in T2/FLAIR volume following steroid increase does not qualify as partial response or stable disease under RANO 2.0 — the corticosteroid qualification rule for response classification applies equally to T2/FLAIR reduction as to enhancement reduction.

8. Practical Radiological Reporting for RANO 2.0

8.1 What Every Glioma Follow-Up Report Must Include

A RANO 2.0-compliant radiology report for a glioma follow-up scan should state:

1. Tumour category: enhancing (Category 1), non-enhancing (Category 2), or mixed (Category 3) — determined by the dominant imaging appearance, not just the molecular type
2. Baseline scan identification: which scan is the RANO 2.0 reference baseline (post-RT scan, date, study ID)
3. Current measurements: SPD of all target lesions (with individual lesion measurements), compared to baseline SPD and nadir SPD
4. Non-target lesion status: qualitative assessment of non-target lesions
5. New lesions: presence or absence; if present — location, size, and whether within or outside radiation field
6. T2/FLAIR assessment: qualitative assessment of change for Category 1; quantitative measurement for Category 2 and 3
7. Corticosteroid change notation: any known change in corticosteroid dose since the last scan, when this information is available
8. Preliminary response classification: CR, MaR, PR, SD, pPD, or PD — with explicit notation of whether the 12-week confirmation window applies

8.2 Measurement Pitfalls to Avoid

Post-surgical cavity: surgical cavities commonly show peripheral rim enhancement due to blood-brain barrier disruption at the resection margin. This must not be confused with tumour enhancement. If the enhancement is clearly following the contour of a surgical cavity and was present on the post-operative baseline scan (prior to radiation), it should not be measured as tumour. New or increasing enhancement at the surgical margin after radiation is more suspicious and should be measured and reported.

Radiation-induced enhancement: enhancement within the radiation field in the first 12 weeks post-RT is characteristically patchy, confluent, or following the radiation dose distribution, rather than the nodular or ring-enhancing morphology of tumour progression. The morphology of enhancement — not just its presence — should be described in the report.

Measuring in the wrong plane: the measurement must be performed in the axial plane for most supratentorial lesions, using the same slice and plane as the baseline measurement. Switching from axial to coronal measurement mid-study introduces artefactual size changes. If the tumour geometry requires a different plane for reliable measurement, this must be declared in the baseline report and maintained throughout.

Pseudoresponse during anti-VEGF therapy: bevacizumab and other anti-angiogenic agents dramatically reduce enhancement without proportional tumour killing. A patient with a clearly shrinking enhancement area while on bevacizumab should not be classified as a partial or major responder unless T2/FLAIR is also controlled — a mixed assessment combining both components (Category 3 logic) applies in this context.

9. Advanced MRI in RANO 2.0: Adjunctive but Unverified

RANO 2.0 explicitly acknowledges that advanced MRI techniques may improve response assessment but states that they require further validation before formal incorporation into the criteria [1]. The specific mention of DSC perfusion MRI as adjunctive for pseudoprogression assessment is a step toward future integration, but does not change the current measurement framework.

The following advanced techniques are used in RANO-adjacent contexts but are not RANO 2.0 criteria:

• DSC perfusion / rCBV: useful for distinguishing pseudoprogression from true progression (see DSC Series, Parts I–III), but not yet a formal RANO criterion
• DWI / ADC: restricted diffusion may indicate hypercellular tumour recurrence; not included in RANO 2.0 measurement framework
• MR spectroscopy: elevated Cho/NAA ratio suggests active tumour; not included
• Amino acid PET (FET, FDOPA, MET): the PET RANO 1.0 criteria were published as a companion document by Lancet Oncology in 2024, providing a parallel framework for PET-based assessment in glioma [8]

The relationship between RANO 2.0 and advanced MRI is expected to evolve as prospective validation data accumulate. The RANO working group explicitly identifies this as an active area of development [1].

10. RANO 2.0 in Clinical Practice vs. Clinical Trials

RANO 2.0 was formally designed for clinical trial use as a standardised, reproducible framework for assessing treatment response endpoints. Its application in routine clinical practice (outside trials) is recommended but not mandatory, and several adaptations are required for real-world implementation:

In clinical practice, full BTIP compliance (parameter-matched 3D IR-GRE pre/post contrast, strict sequence ordering) is the goal but may not always be achievable due to patient tolerance, time constraints, or historical scan incompatibility. The most important RANO-relevant practical elements to implement in routine care are:

• Consistent sequence parameters across serial scans: the single most important technical requirement; if the post-contrast T1 parameters change between examinations, measurements are not comparable
• Identification of the post-RT baseline: every neuro-oncology patient's file should clearly identify the RANO 2.0 baseline scan
• 12-week confirmation window awareness: do not report apparent progression within the radiation field in the first 12 post-RT weeks as confirmed PD in the clinical report
• Consistent measurement approach: 2D SPD at every follow-up; the same plane and the same target lesion selection throughout

The tension between RANO 2.0 and routine practice is real: many radiology departments do not have the workflow, software, or protocol standardisation to achieve full BTIP compliance. The practical minimum — documented baseline, consistent T1 measurement, 12-week confirmation awareness — provides most of the clinical benefit with achievable operational requirements.

11. Evidence-Based References

A. Guidelines / Consensus / Society Recommendations

B. Systematic Reviews / Meta-analyses

C. Important Prospective / Original Studies

D. Technical MRI Papers

E. Landmark Historical References

End of Focus Page — RANO 2.0 Glioma Response Assessment MRI — MRIninja v1.0 — May 2026

Related focus deep dive: Cerebral Gliomas and Glial Tumours — MRI Interpretation.